"test" branch updates

"test" branch updates

[Benjamin Herrenschmidt]

Just a quick note before I leave the office...

I've just pushed out a bunch of stuff to my "test" branch. There's a
couple of known breakage (iseries and 8xx) which will have to be sorted
out before that stuff goes into "next". hopefully in a couple of days.

Still... feel free to take a peek, test, etc...

Cheers,
Ben.

Re: "test" branch updates

[Michael Ellerman]

[-- Attachment #1: Type: text/plain, Size: 874 bytes --]

On Mon, 2009-05-18 at 17:19 +1000, Benjamin Herrenschmidt wrote:
> Just a quick note before I leave the office...
> 
> I've just pushed out a bunch of stuff to my "test" branch. There's a
> couple of known breakage (iseries and 8xx) which will have to be sorted
> out before that stuff goes into "next". hopefully in a couple of days.
> 
> Still... feel free to take a peek, test, etc...

The swiotlb stuff doesn't build:

arch/powerpc/kernel/dma-swiotlb.c
   In function ‘swiotlb_bus_to_virt’:
  29: warning: passing argument 1 of ‘swiotlb_bus_to_phys’ makes integer from pointer without a cast
  29: error: too many arguments to function ‘swiotlb_bus_to_phys’
   At top level:
  42: error: conflicting types for ‘swiotlb_bus_to_phys’


There's changes in linux-next that make it work, which look like they're
going via Ingo?

cheers

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: "test" branch updates

[Kumar Gala]

On Jun 11, 2009, at 9:33 AM, Michael Ellerman wrote:

> On Mon, 2009-05-18 at 17:19 +1000, Benjamin Herrenschmidt wrote:
>> Just a quick note before I leave the office...
>>
>> I've just pushed out a bunch of stuff to my "test" branch. There's a
>> couple of known breakage (iseries and 8xx) which will have to be =20
>> sorted
>> out before that stuff goes into "next". hopefully in a couple of =20
>> days.
>>
>> Still... feel free to take a peek, test, etc...
>
> The swiotlb stuff doesn't build:
>
> arch/powerpc/kernel/dma-swiotlb.c
>   In function =91swiotlb_bus_to_virt=92:
>  29: warning: passing argument 1 of =91swiotlb_bus_to_phys=92 makes =20=

> integer from pointer without a cast
>  29: error: too many arguments to function =91swiotlb_bus_to_phys=92
>   At top level:
>  42: error: conflicting types for =91swiotlb_bus_to_phys=92
>
>
> There's changes in linux-next that make it work, which look like =20
> they're
> going via Ingo?

I think those changes are now in Linus's tree.

- k=